Sensible/latent heat fluxes can be obtained by Large Aperture Scintillometer (LAS) over several kilometers, which play an significant role in the analysis of and application to agricultural and forestry, hydrology and meteorology research. Take LAS observations in Miyun and Guantao stations over Hai River Basin in 2008 as an example. The way of data screening and quality control under unstable conditions as well as the effect of different calculation method on sensible heat flux have been discussed. The results showed: the structure parameter of the refractive index (C2n) should be calculated with the variance of the voltage of structure parameter of the refractive index; Humidity correction can be done with daily Bowen ratio; The effective height of LAS can be calculated with spatial averaged function; The Andreas(1988) function is used to get reliable sensible heat flux. Meanwhile, the nonlinear regression method and dynamic linear regression method have been used to fill the 30min and daily missing data while the 30 min missing data under stable conditions can be set to zero. The correlation relationship between daily ET measured by LAS and EC has been constructed to estimate daily ET when daily Rn<50 W/m2. Based on the above analysis, a set of LAS data processing scheme has been set up, which ensures continuous and high quality sensible/latent data can be obtained over various surfaces and weather conditions.